102 THE FLOWERING PLANT. 
area, the stigma, to which pollen may often be found adhering. 
By splitting open the ovary, several little green ovules may be 
found attached along its upper margin. Now examine a young 
pea-pod, which is a further developed pistil. At one end the 
stalk and calyx will be seen, at the other the remains of the 
style, while the pod is the developed ovary. By holding it up 
to the light, a row of matured ovules, ¢.e., seeds, will be observed 
running along one side. Split open the pod on that side and 
spread it out. It will then look like an ordinary leaf, with well- 
marked midrib and thickened edges bearing ovules. Every 
carpel of an apocarpous pistil is, in fact, regarded as a folded 
leaf, the thickened edges of which bear ovules, and are united 
together in a seam or suture, called the ventral suture. ‘The 
other edge of the carpel may be called the dorsal margin, which 
corresponds to the midrib. The thickened edges here form the 
placenta, and the arrangement or placentation of the ovules is 
here said to be marginal (fig. 48 D). In order to grasp the idea 
that a pistil like that of the pea corresponds to a folded leaf, 
take some simple leaf that tapers gradually to a point, eg., one 
from a fuchsia, and fold it upon the midrib; you will then see 
that the broader part answers to the ovary, and the narrower 
part to the style. It sometimes happens in abnormal flowers 
that carpels remain partly or entirely open, thus showing their 
true nature, and the syncarpous pistil of mignonette never com- 
pletely closes at the top (cf also p. 107). In gymnosperms 
folding has not taken place at all. We may consider that the 
pea has descended from ancestors somewhat resembling the 
gymnosperms of the present day, and that the formation of a 
closed ovary has taken place gradually in the course of innu- 
merable generations. The next question is, “‘ Which surface of 
the folded leaf is inside, upper or lower?” ‘This query cannot 
easily be answered by reference to the pea alone, but examination 
of apocarpous pistils with more than one carpel will readily 
give a solution. Examine, for instance, a head of ripe fruit 
in larkspur, columbine, or marsh-marigold, where three, five, 
and several carpels are respectively present. It will be seen 
that the ventral sutures face inwards, for in the fruit they will 
have split open, allowing the seeds to be seen attached to their 
edges. But facing inwards means facing the shortened axis or 
receptacle, and since folding has taken place in this direction, 
the wpper surface of the leaf must bound the internal cavity. 
Take a piece of stem with a foliage leaf attached, and fold this 
up so that the approximated edges face the stem, and the nature 
of the above process will be seen. In the single carpel of pea 
the ventral suture faces upwards ; and since there are five sepals, 
